Thursday 31 December 2009

New Year's Eve

'THIS IS THE END OF THE BEGINNING': Final image from George Pal's 'Destination Moon' (1950)Okay, that's it. First decade of the new century over, and we've got almost nothing good to show for it, physics-wise.
That's bad. We only get ten of these per century. One down, only another nine to go before 2100. If we're burning through resources at the current rate, we can't afford to waste decades like this if we want to actually achieve something significant this century before we get hit by a resources crash.

So a suggested schedule. Let's officially notice the idea of a no-floor implementation of GR by at least late 2010, and see if we can get rid of dark matter and dark energy. Let's have the quantum gravity guys working on acoustic metrics as a low-velocity approximation have the guts to come out and actually suggest that this might be the basis of a real theory, and not just a toy model. Let's stop issuing press releases claiming that the current version of general relativity is the wonderfullest theory and has never ever failed us, let's acknowledge the problems and let's sit down and write a proper general theory from scratch, stealing that "acoustic metric" work.

Instead of setting a schedule that puts the next theoretical breakthroughs maybe eighty or a hundred years from now because we aren't clever enough to understand string theory, let's get off our arses and do the things that we do know how to do. Kick off with the no-floor approach, and when we're energised by the success of that, converge the acoustic metric work with a GR rewrite .. and suddenly the next generation of theory only looks about five years away. If we're very lucky, two and a half. If we can't get enough people onboard fast enough, maybe eight to ten.

Unless we take that first step of exploring the idea that change might be possible and might be a good thing, we won't get anywhere except by dumb luck and/or massive public spending on hardware. If we're not careful, and we don't change the way we do things, next thing we know it'll be 2020 and we still won't have achieved anything.

So let's write off the 00's as a big double-zero. Let's pretend that the Bush years and Iraq and the financial crash never happened. We don't need multi-billion-dollar hardware for this, we only need to be able to think, and to be a bit more adventurous than we've been for the last few decades. Lets redo general relativity properly and get a theory that we can be proud of without having to spin results, one that actually predicts new effects in advance rather than retrospectively, and has the potential to lead us into genuinely new physics territory.

Tomorrow is 2010. Let's start again.

1 comment:

Oliver Sparrow said...

On barking up unclimable trees (= wrong trees).

I like your attraction to thought experiments. IMHO the phenomenology introduced into physics 1930-50, ff, reduced our ability to think.

We are, however, left with a grand theory of everything-shaped hole. It is interestig to think what coudl fill such a hole: not its content, but its architecture.

First, it has to self-assemble from some abstract verity or verities: e.g. cardinality of numbers, sphere packing in high dimension space. These can have nothing to do with actual observables, as those have to emerge from the underlying entities.

One thing that has to be constructed is a spacetime, with assorted broken symmetries that look like e.g. fields. Space is a housekeeping deveice to account for event sequencing, time is mysterious. The vectro product of anything in spacetiem has, however, hae to propagate at light speed, and revolve to be more space like and less time like in response to interactions from other "anythings", breaks in its symmetry.

That is quite a tall order. However, string theory seems to build its stage sets out of other, even more elaborate stage sets and so on ad infinitum

An alternative thought is this. Emergence coems from theincreased complexity that interactiosn of relatively simple entities can geenrate. We note it when the model that describes any exemplar of these entities fails to deal with the composite: pahse changes and molecules, market prices and individual shoppers.

Symbolise the emergce change of pace as A > B, where B consitiutes the emergent and A the source.

Now:
A>B
B>C
C>A
(Of course, this is a toy ring: one coudl have much more complex networks.) The point is that a bootstrap is possible, and so is (infinite) complexity. But. I think you need T before any of this can happen; if not the T we know and love, then something that conserves sequentiality and affords action, like a chip clock.